| { | |
| "best_metric": 0.909375, | |
| "best_model_checkpoint": "vit-large-patch16-224-finetuned-landscape-test/checkpoint-270", | |
| "epoch": 30.0, | |
| "eval_steps": 500, | |
| "global_step": 300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.163173198699951, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 1.3906, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.496875, | |
| "eval_loss": 1.1521408557891846, | |
| "eval_runtime": 6.5609, | |
| "eval_samples_per_second": 48.774, | |
| "eval_steps_per_second": 1.524, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 5.007761478424072, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.914, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.66875, | |
| "eval_loss": 0.781201183795929, | |
| "eval_runtime": 6.7058, | |
| "eval_samples_per_second": 47.72, | |
| "eval_steps_per_second": 1.491, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 6.4083170890808105, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6704, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.76875, | |
| "eval_loss": 0.556643009185791, | |
| "eval_runtime": 6.6958, | |
| "eval_samples_per_second": 47.791, | |
| "eval_steps_per_second": 1.493, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 2.2590925693511963, | |
| "learning_rate": 4.814814814814815e-05, | |
| "loss": 0.4609, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.83125, | |
| "eval_loss": 0.4362529218196869, | |
| "eval_runtime": 6.6327, | |
| "eval_samples_per_second": 48.245, | |
| "eval_steps_per_second": 1.508, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.843147039413452, | |
| "learning_rate": 4.62962962962963e-05, | |
| "loss": 0.404, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.815625, | |
| "eval_loss": 0.4806601405143738, | |
| "eval_runtime": 6.7484, | |
| "eval_samples_per_second": 47.418, | |
| "eval_steps_per_second": 1.482, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 3.1339221000671387, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.3948, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.853125, | |
| "eval_loss": 0.4215615689754486, | |
| "eval_runtime": 6.7585, | |
| "eval_samples_per_second": 47.348, | |
| "eval_steps_per_second": 1.48, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 1.822701096534729, | |
| "learning_rate": 4.259259259259259e-05, | |
| "loss": 0.3535, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.86875, | |
| "eval_loss": 0.32805871963500977, | |
| "eval_runtime": 6.7133, | |
| "eval_samples_per_second": 47.666, | |
| "eval_steps_per_second": 1.49, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 2.6920511722564697, | |
| "learning_rate": 4.074074074074074e-05, | |
| "loss": 0.3107, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9, | |
| "eval_loss": 0.2971627116203308, | |
| "eval_runtime": 6.6598, | |
| "eval_samples_per_second": 48.049, | |
| "eval_steps_per_second": 1.502, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 3.074315309524536, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 0.3086, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.88125, | |
| "eval_loss": 0.332777202129364, | |
| "eval_runtime": 6.6439, | |
| "eval_samples_per_second": 48.165, | |
| "eval_steps_per_second": 1.505, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 1.8536041975021362, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 0.2564, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8875, | |
| "eval_loss": 0.35169774293899536, | |
| "eval_runtime": 6.7081, | |
| "eval_samples_per_second": 47.704, | |
| "eval_steps_per_second": 1.491, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 2.0643813610076904, | |
| "learning_rate": 3.518518518518519e-05, | |
| "loss": 0.2654, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.859375, | |
| "eval_loss": 0.39852675795555115, | |
| "eval_runtime": 6.6638, | |
| "eval_samples_per_second": 48.021, | |
| "eval_steps_per_second": 1.501, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 2.567340135574341, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.2733, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.90625, | |
| "eval_loss": 0.28696078062057495, | |
| "eval_runtime": 6.66, | |
| "eval_samples_per_second": 48.048, | |
| "eval_steps_per_second": 1.501, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 2.107881546020508, | |
| "learning_rate": 3.148148148148148e-05, | |
| "loss": 0.2511, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.8875, | |
| "eval_loss": 0.41765227913856506, | |
| "eval_runtime": 6.7346, | |
| "eval_samples_per_second": 47.516, | |
| "eval_steps_per_second": 1.485, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 2.686899423599243, | |
| "learning_rate": 2.962962962962963e-05, | |
| "loss": 0.2762, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.89375, | |
| "eval_loss": 0.35794347524642944, | |
| "eval_runtime": 6.6606, | |
| "eval_samples_per_second": 48.044, | |
| "eval_steps_per_second": 1.501, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 1.5542501211166382, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.2188, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.890625, | |
| "eval_loss": 0.33477360010147095, | |
| "eval_runtime": 6.6577, | |
| "eval_samples_per_second": 48.064, | |
| "eval_steps_per_second": 1.502, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 1.8425265550613403, | |
| "learning_rate": 2.5925925925925925e-05, | |
| "loss": 0.2265, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.903125, | |
| "eval_loss": 0.3045758605003357, | |
| "eval_runtime": 6.6513, | |
| "eval_samples_per_second": 48.111, | |
| "eval_steps_per_second": 1.503, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 1.9955267906188965, | |
| "learning_rate": 2.4074074074074074e-05, | |
| "loss": 0.2054, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.896875, | |
| "eval_loss": 0.33053359389305115, | |
| "eval_runtime": 6.639, | |
| "eval_samples_per_second": 48.2, | |
| "eval_steps_per_second": 1.506, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 2.128180980682373, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.1951, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.88125, | |
| "eval_loss": 0.3575904369354248, | |
| "eval_runtime": 6.6576, | |
| "eval_samples_per_second": 48.065, | |
| "eval_steps_per_second": 1.502, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 1.225751280784607, | |
| "learning_rate": 2.037037037037037e-05, | |
| "loss": 0.1762, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.88125, | |
| "eval_loss": 0.39847835898399353, | |
| "eval_runtime": 6.6313, | |
| "eval_samples_per_second": 48.256, | |
| "eval_steps_per_second": 1.508, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 2.6766974925994873, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 0.2264, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.903125, | |
| "eval_loss": 0.3710865080356598, | |
| "eval_runtime": 6.7413, | |
| "eval_samples_per_second": 47.469, | |
| "eval_steps_per_second": 1.483, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "grad_norm": 1.814340591430664, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.1958, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_accuracy": 0.8875, | |
| "eval_loss": 0.32590168714523315, | |
| "eval_runtime": 6.6661, | |
| "eval_samples_per_second": 48.004, | |
| "eval_steps_per_second": 1.5, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "grad_norm": 1.8976715803146362, | |
| "learning_rate": 1.4814814814814815e-05, | |
| "loss": 0.1765, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_accuracy": 0.89375, | |
| "eval_loss": 0.380399614572525, | |
| "eval_runtime": 6.7489, | |
| "eval_samples_per_second": 47.415, | |
| "eval_steps_per_second": 1.482, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "grad_norm": 2.2366738319396973, | |
| "learning_rate": 1.2962962962962962e-05, | |
| "loss": 0.1859, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_accuracy": 0.9, | |
| "eval_loss": 0.34643369913101196, | |
| "eval_runtime": 6.6848, | |
| "eval_samples_per_second": 47.87, | |
| "eval_steps_per_second": 1.496, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 1.6862928867340088, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 0.1915, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_accuracy": 0.890625, | |
| "eval_loss": 0.3741697371006012, | |
| "eval_runtime": 6.6221, | |
| "eval_samples_per_second": 48.323, | |
| "eval_steps_per_second": 1.51, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 1.770437240600586, | |
| "learning_rate": 9.259259259259259e-06, | |
| "loss": 0.1667, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_accuracy": 0.90625, | |
| "eval_loss": 0.3199695944786072, | |
| "eval_runtime": 6.6194, | |
| "eval_samples_per_second": 48.343, | |
| "eval_steps_per_second": 1.511, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "grad_norm": 1.6922975778579712, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 0.1744, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_accuracy": 0.89375, | |
| "eval_loss": 0.3544948101043701, | |
| "eval_runtime": 6.6657, | |
| "eval_samples_per_second": 48.007, | |
| "eval_steps_per_second": 1.5, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "grad_norm": 1.4301420450210571, | |
| "learning_rate": 5.555555555555556e-06, | |
| "loss": 0.1595, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "eval_accuracy": 0.909375, | |
| "eval_loss": 0.3100855350494385, | |
| "eval_runtime": 6.6671, | |
| "eval_samples_per_second": 47.997, | |
| "eval_steps_per_second": 1.5, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "grad_norm": 1.7006909847259521, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 0.1793, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_accuracy": 0.896875, | |
| "eval_loss": 0.3230052888393402, | |
| "eval_runtime": 6.6338, | |
| "eval_samples_per_second": 48.238, | |
| "eval_steps_per_second": 1.507, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "grad_norm": 2.8452911376953125, | |
| "learning_rate": 1.8518518518518519e-06, | |
| "loss": 0.1596, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "eval_accuracy": 0.9, | |
| "eval_loss": 0.32675567269325256, | |
| "eval_runtime": 6.6591, | |
| "eval_samples_per_second": 48.054, | |
| "eval_steps_per_second": 1.502, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 2.136308431625366, | |
| "learning_rate": 0.0, | |
| "loss": 0.169, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_accuracy": 0.896875, | |
| "eval_loss": 0.3321378529071808, | |
| "eval_runtime": 6.6284, | |
| "eval_samples_per_second": 48.277, | |
| "eval_steps_per_second": 1.509, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "step": 300, | |
| "total_flos": 1.0519143604184678e+19, | |
| "train_loss": 0.31787962436676026, | |
| "train_runtime": 2532.1955, | |
| "train_samples_per_second": 15.165, | |
| "train_steps_per_second": 0.118 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 300, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.0519143604184678e+19, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |